Date of the Workshop: Tuesday 30 June 2026
Location: University of Southampton
Number of participants: Approximately 100 (a core of about 50 were there for all sessions)
Number of distinct organisations represented: 3 Universities. Multiple departments.
Introduction
As generative artificial intelligence reshapes how students learn, conduct research, and develop analytical skills, Higher Education Institutions are re-evaluating their teaching and assessment strategies. These debates are particularly pressing in quantitative fields.
In this workshop, we examine how generative AI can be integrated as a learning and research companion across the dissertation process within quantitative disciplines. Particular attention will be given to how generative AI can help students strengthen core competencies which are highly valued in the job market, including quantitative analysis, critical thinking, and data literacy. At the same time, we will reflect on the skills that remain fundamentally human, such as the ability to communicate research effectively in interdisciplinary settings, and how these shape not only how we teach, but also, indirectly, WHAT we teach. While we will briefly acknowledge the broader implications for curriculum design, the workshop primarily focuses on HOW AI can practically support learning in quantitative subjects.
The workshop was designed to be relevant to academics across disciplines to enhance support for students at different stages of their programme and highlights how AI can enhance both learning and research practices.
Workshop Overview
The workshop was in-person but we had the capability to have external attendees.
Morning Session: Antonio Mele (LSE)
AI Integration in Research and Teaching: Antonio led a comprehensive session, detailing how AI—particularly agentic AI—can be integrated into research workflows, student supervision, and teaching, highlighting practical applications, challenges, and institutional considerations at LSE. This took the form of a demonstration/lecture followed by questions. This was followed up by a late afternoon workshop based on this introduction.
Afternoon Session 1: (20 min talks/discussions based on educators use of AI in Education Activities)
Internal Contributors (University of Southampton): Dr Monica Breeder (School of Economics, Social and Political Sciences) , Dr Ben Davis (Southampton Education School), Prof. Tom Ezard (Associate Dean (Education), Faculty for Environmental and Life Sciences)
External Contributor: Assoc. Prof. Alun Owen (Assoc. Professor in Statistics & Assoc. Director of sigma Maths & Stats Support, Coventry University)
Teaching Students to Think Critically with AI: Monica gave a talk and then led a discussion on the necessity of teaching students to use AI and large language models (LLMs) critically, focusing on economics education and the challenges of integrating AI into student research and assessment, while highlighting the importance of critical thinking, reference verification, and AI literacy for future employability.
Agentic AI, STACK, and the changing dynamics of question authoring: Ben presented on the use of AI, specifically multi-agent frameworks and large language models, to digitise and scale high-quality mathematics textbook content and automate the creation of interactive assessment questions, sharing practical lessons from implementing these systems in educational settings.
Integrating Generative AI into Biological Data Science Education: Tom discussed the integration of generative AI into a biological data science module, detailing the evolution of assessment design to foster critical analysis of AI-generated code and outputs, and the challenges of balancing open-ended tasks with the need for structured, meaningful evaluation.
Developing a Generative AI Agent to support HE students’ learning in statistics for dissertations: Alan presented the design and deployment of ‘Sigma Chat’, a specialised AI agent tailored to support students with statistics, particularly those with low confidence or from non-specialist backgrounds, and compared its performance to standard generative AI tools.
Afternoon Session 2: Workshop (run by Antonio Mele, based on the material from the morning session)
Technical Workshop: Coding Agents, Project Governance, and Automation in Research: The latter part of the meeting, led by Emanuela and others, provided a technical workshop on the use of coding agents, project governance files, skills, sub-agents, and automation hooks in research workflows, with practical demonstrations and guidance on best practices for managing AI-driven research projects
Note: throughout the day there was space to connect with other educators.
Main Themes and Ideas
Broadly: Agentic AI Integration in Research and Teaching – extended seminar and workshop.
AI in educational practice: AI and Critical thinking, Assessment Design (in the age of AI and using AI), Critical Assessment of AI developed code, AI Tutor development
Emerging Implications and Recommendations
The two big topics from discussions outside of the talks:
- How Education is/will be changing:
What skills do our graduates need/expect
How do we ensure a worthwhile student experience
How do we ensure rigor in our programmes
How do we ensure QA for our networks
Local level workshops are needed to give colleagues time, space, thought and support as to how they are going to manage the changing face of Education (undergraduate, PGT and PGR).
Collaborations with industry (highlighting use, critical thought) to make sure we are current with respect to industry standards of AI acceptable AI use.
- How does the Education Framework change
How do we talk to VCs/PVCs/OfS/Government about changes, new ways of thinking
How do we engage with policy makers (from upper echelons in OfS, to policymakers etc)
In terms of IMA/RSS/LMS workshops. These local ones were great. I could not attend the others (though I did instigate meetings with colleagues at UCL so we didn’t overlap too much… however I could not attend their meeting). Maybe having a similar grant call next year but with the caveat that the applications must include multiple institutions and a proposed strand for collaboration.
Some “state of the sector” sessions whereby policymakers talk through (and ask for debate/discussion) the proposed direction of travel etc.
More case-studies from staff that can be online sessions will all disciplines able to highlight interesting work and possibilities for use across disciplines.
Resources and Links
https://sotonac.sharepoint.com/teams/FSSADEducation/SitePages/AI-Workshop-30-June-2026.aspx
Acknowledgments
Thanks to the IMA/RSS/LMS for part funding the workshop. Additional funding came from my Associate Dean (Education) budget. So additional thanks to the Faculty of Social Sciences for that. Thanks to Antonio and Alun for travelling down and sharing their experience and insight and to Monica, Ben and Tom for talking through how they are engaging students with AI and using AI themselves.
A big thank you to Prof. Emanuela Lotti (Economics) who was the co-organiser with me and to Jana Sedah, Thomas Gall (both Economics), Ruben Sanchez-Garcia (Maths), Athina Toma and Ben Davies (Education) for their support in reading the application and with the structure of the day.
Alex Silvestre made sure the day ran smoothly.
Conclusion
We are already planning our next set of workshops. The first one will be about AI marking of formative work and we are hoping to plan a series of monthly sessions for the Faculty and wider University Community.
